THS4503 Tina Problem

Jeremy Partain1

Other Parts Discussed in Thread: THS4503, THS4222, THS4631

When I implement a single ended to differential circuit using the THS4503 by itself (no other amplifiers, only passive components/voltage generators/batteries), the simulation results seem to be accurate. When I add a THS4222 amp to feed the THS4503 instead of a voltage generator, the difference output of the THS4503 is 0v (if I zoom way in, I can see the signal is only 100mv instead of the expected 5v).
I have verified the output of the THS4222 in the second simulation is the same as the output from the voltage generator from the first simulation - ie the input to both THS4503 circuits is the same.

Can Tina get confused when multiple components are added to a model? I will upload a picture of the model and the result output soon.

over 8 years ago

0 Jeremy Partain1 over 8 years ago

Prodigy 130 points

Here is the THS4503 simulation by itself:

And here is both the THS4222 and THS4503 together:

I cant figure out why the output of the THS4503 is almost non existent. I know there is also a THS4631, but the problem exists any time there is any other amplifier besides the THS4503 in the circuit.

0 Samir Cherian over 8 years ago in reply to Jeremy Partain1

TI__Genius 12631 points

Hi Jeremy,
In Case 1 your input signal (Vinput) from the generator is sitting on a 2.5V dc offset. In case 2, you are sitting on a dc offset of 2.0V. Can you please change R2 in circuit 2 to 3k, which makes it a 1/4 divider and resimulate your circuit. If that doesnt work, can you please attach your TINA circuits so I can play around with it.
Thanks,samir

0 Jeremy Partain1 over 8 years ago in reply to Samir Cherian

Prodigy 130 points

TSH4503.TSC

THS4631.TSC

Sorry for the bad screen shot. Everything is connected correctly, but the screen shot is poor quality so some of the connections have disappeared. Thanks for looking at this.. Attached are the two models and the output after I made the fix you asked for. The problem is that for nearly the same Vinput to the THS4503, the output goes from realistic in the first model when the THS4503 is the only amp in the circuit, to almost 0v when any other amplifier is in the circuit with it.

0 Samir Cherian over 8 years ago in reply to Jeremy Partain1

TI__Genius 12631 points

Hi Jeremy,
I tried your circuits and it does indeed look like an issue with the THS4503 model. I will forward this over to the SPICE team for rectification. Thanks for bringing this to our attention and dorry for the inconvenience. I will let you know once I hear back from the team.
-Samir

0 Jeremy Partain1 over 8 years ago in reply to Samir Cherian

Prodigy 130 points

In the mean time, is there any other tool I can use to model my circuit?

0 Samir Cherian over 8 years ago in reply to Jeremy Partain1

TI__Genius 12631 points

Hi Jeremy, Have you tried using split supplies? When I tried it the other day with bipolar supplies, the issue seemed to go away.
Thanks,
samir

0 Jeremy Partain1 over 8 years ago in reply to Samir Cherian

Prodigy 130 points

I just added a -5v supply onto the THS4503 amp and the output remained nulled. Is that where you added the negative supply?

0 Samir Cherian over 8 years ago in reply to Jeremy Partain1

TI__Genius 12631 points

Hi Jeremy,

Please see attached schematic. You need to change between 2 configurations to see the difference in performance:

Config 1 - Single Supply: In this configuration the device gain is incorrect.

Vdd= 5V

Vee = 0V

Vcm = 2.5V

Vg1 (dc offset) = 2.5V

Config 2 - Split supplies: In this configuration the device gain is correct.

Vdd= 2.5V

Vee = -2.5V

Vcm = 0V

Vg1 (dc offset) = 0V

Let me know if this works.

-Samir

THS4503_Issue.TSC

0 Jeremy Partain1 over 8 years ago in reply to Samir Cherian

Prodigy 130 points

I took your model and changed the parameters to be like Config 1 from your last post and I got the following nulled output:

However, when I simply removed the THS4222, I get the following output which is correct:

The only thing changed in the simulation was removing the THS4222. This is why I think there is a bug in the THS4503 spice model.

Does this prove the model is broken?

0 Samir Cherian over 8 years ago in reply to Jeremy Partain1

TI__Genius 12631 points

The model is definitely broken. The modeling team is looking into it. They believe there is a floating node somewhere in the macro-model.

0 Jeremy Partain1 over 8 years ago in reply to Samir Cherian

Prodigy 130 points

Did the sim team ever get the THS4503 model fixed?

0 Samir Cherian over 8 years ago in reply to Jeremy Partain1

TI__Genius 12631 points

Hello Jeremy,
The sim team looked into this and they found a potential fix...what you need to do in TINA is :
1. On the top menu, click on Analysis -> Set Analysis Paramaters...
2. Once you are there look for line item "Always start with source step". You need to make this Yes. The simulation seems to work properly after that. You may have to click on the Index finger button at the bottom of the menu, and then hit "View All" in order to see all the available parameters.
-Samir

Amplifiers

Amplifiers forum

THS4503 Tina Problem